Home A Workflow to Improve the Alignment of Prostate Imaging with Whole-mount Histopathology
Post
Cancel

A Workflow to Improve the Alignment of Prostate Imaging with Whole-mount Histopathology

Posted Jul 31, 2014 Updated Apr 10, 2023
By Hidekazu Yamamoto MB BChir MRCS PhD , Dror Nir PhD , Lona Vyas MB BS MRCS , Richard T. Chang MB BS MRCS , Rick Popert MS FRCS (Urol) , Declan Cahill MSc FRCS (Urol) , Ben Challacombe BSc (Hons) MS FRCS (Urol) , Prokar Dasgupta MSc (Urol) DLS MD FRCS FRCS (Urol) FEBU , Ashish Chra FRCPath DipRCPath (Cytol)
12 min read

Rationale and Objectives

Evaluation of prostate imaging tests against whole-mount histology specimens requires accurate alignment between radiologic and histologic data sets. Misalignment results in false-positive and -negative zones as assessed by imaging. We describe a workflow for three-dimensional alignment of prostate imaging data against whole-mount prostatectomy reference specimens and assess its performance against a standard workflow.

Materials and Methods

Ethical approval was granted. Patients underwent motorized transrectal ultrasound (Prostate Histoscanning) to generate a three-dimensional image of the prostate before radical prostatectomy. The test workflow incorporated steps for axial alignment between imaging and histology, size adjustments following formalin fixation, and use of custom-made parallel cutters and digital caliper instruments. The control workflow comprised freehand cutting and assumed homogeneous block thicknesses at the same relative angles between pathology and imaging sections.

Results

Thirty radical prostatectomy specimens were histologically and radiologically processed, either by an alignment-optimized workflow ( n = 20) or a control workflow ( n = 10). The optimized workflow generated tissue blocks of heterogeneous thicknesses but with no significant drifting in the cutting plane. The control workflow resulted in significantly nonparallel blocks, accurately matching only one out of four histology blocks to their respective imaging data. The image-to-histology alignment accuracy was 20% greater in the optimized workflow ( P < .0001), with higher sensitivity (85% vs. 69%) and specificity (94% vs. 73%) for margin prediction in a 5 × 5-mm grid analysis.

Conclusions

A significantly better alignment was observed in the optimized workflow. Evaluation of prostate imaging biomarkers using whole-mount histology references should include a test-to-reference spatial alignment workflow.

Measuring the accuracy of imaging biomarkers to localize prostate cancer is a complex task that involves correlating the match in the zonal distribution of lesions between the test imaging data and histopathologic reference. In most studies that use whole-mount radical prostatectomy specimens as “gold standard” references , important assumptions are that histologic and radiologic zonal boundaries are aligned to each other and that zonal assignment of lesions is accurate ( Fig 1 ). Such assumptions, however, would not hold if zonal boundaries are misaligned ( Fig 1 a,b), or histologic sectioning is variable ( Fig 1 c), both of which will reduce the overall accuracy results and undermine the internal validity of the study as set out in the Standards for the Diagnostic Accuracy Studies guidelines . There are no validated methods to assess the spatial alignment between imaging and histologic data and to ensure that misalignment is minimized.

Open full size image

Figure 1

Problems associated with zonal correlation of lesions between whole-mount pathology with radiology data are shown schematically in sagittal illustrations of the prostate. Lesions are indicated as a blue spot . Blue or red lines indicate boundaries of zones or slices, respectively. In this example of a three-zone analysis, a lesion can lie either in the midzone alone or both to the midzone and apex, depending on how zonal boundaries are defined (a) . Hence, an imaging biomarker with 100% accuracy can still result in poor accuracy if zonal boundaries between pathology and imaging are not aligned (a) . In comparing step sections of pathology and radiology, differences in relative angles of sectioning lead to a similar result (b) . Errors in zonal assignment can also arise from incorrect assumptions regarding the quality of sectioning (c) . Lesions are usually assigned to zones by interpolating the findings of step sections. Hence, a lesion involving sections 2 and 3 can be localized to the midzone, if sections were all cut at a known thickness and in a plane perpendicular to the apical-basal axis. However, if sections were cut in parallel but with heterogeneous thicknesses, or in a nonparallel manner, assumptions of equal and parallel sections could result in lesions being mislocalized. (Color version of figure is available online.)

Get Radiology Tree app to read full this article<

Materials and methods

Patients

Get Radiology Tree app to read full this article<

Alignment-optimized Workflow

Get Radiology Tree app to read full this article<

Figure 2, The workflow optimized for spatial alignment of radiologic test imaging to histopathologic reference data is shown. Pathology and radiology workflows are equivalent in definitions of axes, thus allowing the measurement of relative angles and positions between histology and pathology sections. Cutting of tissue sections is assisted by a tissue planer, and accuracy of cutting is assessed by thickness measurements at multiple points per section using calipers. Adjustments are made to account for dimensional changes of the prostate due to formalin fixation. The pathology and radiology workflows converge for image registration, which are composed of two alignment and two transformation steps. Once histology sections are matched to corresponding imaging sections, it is possible to progress to subsequent zonal analyses or a mapping analysis of cancer distribution. AB, apical-basal; AP, anterior-posterior; FFC, formalin fixation coefficient; MRI, magnetic resonance imaging; T, transverse. (Color version of figure is available online.)

Get Radiology Tree app to read full this article<

Imaging Analysis

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Formalin Fixation and Recording of Macroscopic Descriptions

Get Radiology Tree app to read full this article<

Figure 3, (a) A probe is inserted into the urethra to mark the apical-basal axis. Different colored inks are applied to each side of the prostate. (b) Seminal vesicles are amputated. (c) Dimensions of the formalin-fixed prostate are recorded in the apical-basal, anterior-posterior, and transverse axes.

Get Radiology Tree app to read full this article<

Formalin Fixation Coefficients

Get Radiology Tree app to read full this article<

Generation of Tissue Sections

Get Radiology Tree app to read full this article<

Figure 4, A sketch of the tissue cutting device is shown (a) . The formalin-fixed prostate is placed on the space marked “X” on the device with the posterior surface facing down. A probe in the urethra is used to align the anterior-posterior axis with the marked midline on the device. The base of the gland was gently pressed onto surface “Y”. The probe is subsequently removed, and a mounted microtome blade lowered along the 4-mm raised edge of the device from top to bottom to cut each block (b) . Sliced blocks were put into tissue-processing cassettes with the apical face facing down. The cutting process was repeated for the remainder of the prostate (c) . For each block, thicknesses were measured at five different sectors (anterior, posterior, left, right, and center, each marked “X”) (d) . (Color version of figure is available online.)

Get Radiology Tree app to read full this article<

Assessment of Cutting Error Using the Tissue Planer

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Examination and Scanning of Pathologic Slides

Get Radiology Tree app to read full this article<

Alignment and Matching of Pathologic Slides to Imaging Blocks in the AB Axis

Get Radiology Tree app to read full this article<

Figure 5, The process of matching pathology slides to corresponding imaging blocks is shown in sagittal views of the prostate. Thicknesses of tissue blocks ( orange arrows ) were used to determine the slide coordinates along the AB axis of the prostate. Histology slides were cut from the apical surfaces of tissue blocks 1–3, marked as A–C, respectively. Imaging blocks corresponding to each slide were identified using AB axis coordinates of each slide ( red arrowed line ). As shown in this example, slides can be matched to different imaging blocks depending on whether FFC adjustment was used. Without FFC adjustment, slides were matched to imaging blocks 1, 2, and 3, respectively. With FFC adjustment in the AB axis ( blue arrow ), slides A, B, and C were matched to imaging blocks 2, 3, and 4, respectively. AB, apical-basal; FFC, formalin fixation coefficient. (Color version of figure is available online.)

Get Radiology Tree app to read full this article<

Alignment and Assessment of Overlay Accuracy between Pathologic Slides and Imaging Blocks in the Transverse Plane

Get Radiology Tree app to read full this article<

Figure 6, The process of matching pathology slides to corresponding imaging blocks in the transverse plane and subsequent assessment of alignment are shown. Scanned images of histopathology slides were resized by linear transformation to their prefixation size by multiplying the length and width of the image by their FFCs in the AP and T axes, respectively (a,b) . Corresponding PHS transverse projection images were also resized to their original sizes using length (a,b) . A 5 × 5-mm grid was overlaid onto all images, with the midpoint of the grid defined as the midpoints of the AP and T axes (c) . Alignment accuracy between radiology and pathology images was assessed by the alignment of margins between the overlaid images, scoring each square within the overlaid grids as either “1” indicating the presence of a histologic or radiologic margin or “0”” for no margin. Scored pathology grids were used as the reference, and scored radiology grids were used as the index. In this example, we show the numbers of true positives ( yellow squares , n = 25), false positives ( green squares , n = 4), false negatives ( red squares , n = 3), and true negatives ( gray squares , n = 38). AP, anterior-posterior; FFC, formalin fixation coefficient; PHS, Prostate Histoscanning; T, transverse. (Color version of figure is available online.)

Get Radiology Tree app to read full this article<

Statistical Analysis

Get Radiology Tree app to read full this article<

Results

The Effect of Formalin Fixation on Prostate Dimensions and Volume

Get Radiology Tree app to read full this article<

Figure 7, (a) The linear regression line shows the relation between pathologic weight and ultrasound volume segmented by Prostate Histoscanning ( r 2 = 0.81). (b) Linear regression lines show the relation between prostate dimensions measured by transrectal ultrasound and by calipers. Path, pathology; US, ultrasound. (Color version of figure is available online.)

Get Radiology Tree app to read full this article<

Parallel Sectioning Accuracy of the Tissue Planer Versus Freehand Cutting

Get Radiology Tree app to read full this article<

Figure 8, (a) A dot plot of block thicknesses is shown for specimens cut using either the tissue planer (cases 1–20) or freehand technique (cases A–K). Each dot represents the thickness of one block (horizontal bar + mean thickness). The tissue planer generated thinner cuts compared to freehand cutting (4.58 vs. 5.69 mm, P < .0001, unpaired t test), with higher precision within patients (0.55 vs. 1.14 mm) but less precision between patients (0.80 vs. 0.39 mm). (b) Totals of block sector thicknesses ( Fig 4 d) were calculated for each patient and compared between anterior and posterior sectors and between left and right sectors, to determine the drift in the cutting plane with sequential cutting using either the tissue planer or freehand technique. The difference was significantly different from zero only in the freehand group comparing of anterior and posterior sectors (95% confidence intervals, −1.8 to −6.6 mm).

Get Radiology Tree app to read full this article<

Assessment of Alignment in the AB Axis

Get Radiology Tree app to read full this article<

Alignment in the Transverse Plane

Get Radiology Tree app to read full this article<

Table 1

Contingency Tables Comparing Margin Overlay Accuracy between Optimized and Control Alignment Workflows

Radiology, Margin Present Radiology, No Margin Optimized workflow—prostate sectioning by tissue planer, with alignment based on FFC-adjusted coordinates in all axes Histology, margin Present 1692 307 Histology, no margin 256 3737 Control workflow—freehand prostate sectioning, with alignment without FFC-adjusted coordinates Histology, margin present 571 256 Histology, no margin 526 1425

Get Radiology Tree app to read full this article<

Discussion

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Conclusions

Get Radiology Tree app to read full this article<

Acknowledgments

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

References

  • 1. Rosenkrantz A.B., Mannelli L., Kong X., et. al.: Prostate cancer: utility of fusion of T2-weighted and high b-value diffusion-weighted images for peripheral zone tumor detection and localization. J Magn Reson Imaging 2011; 34: pp. 95-100.

  • 2. Jackson A.S., Reinsberg S.A., Sohaib S.A., et. al.: Dynamic contrast-enhanced MRI for prostate cancer localization. Br J Radiol 2009; 82: pp. 148-156.

  • 3. Hambrock T., Somford D.M., Huisman H.J., et. al.: Relationship between apparent diffusion coefficients at 3.0-T MR imaging and Gleason grade in peripheral zone prostate cancer. Radiology 2011; 259: pp. 453-461.

  • 4. Delongchamps N.B., Rouanne M., Flam T., et. al.: Multiparametric magnetic resonance imaging for the detection and localization of prostate cancer: combination of T2-weighted, dynamic contrast-enhanced and diffusion-weighted imaging. BJU Int 2011; 107: pp. 1411-1418.

  • 5. Vargas H.A., Akin O., Franiel T., et. al.: Diffusion-weighted endorectal MR imaging at 3 T for prostate cancer: tumor detection and assessment of aggressiveness. Radiology 2011; 259: pp. 775-784.

  • 6. Bossuyt P.M., Reitsma J.B., Bruns D.E., et. al.: Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. BMJ 2003; 326: pp. 41-44.

  • 7. Ashburner J., Fristo K.J.: Chapter 2, “Rigid Body Registration”. Human brain function.2 ed.2004.Academic Press

  • 8. Salomon G., Spethmann J., Beckmann A., et. al.: Accuracy of HistoScanning for the prediction of a negative surgical margin in patients undergoing radical prostatectomy. BJU Int 2013; 111: pp. 60-66.

  • 9. Simmons L.A., Autier P., Zat’ura F., et. al.: Detection, localisation and characterisation of prostate cancer by prostate HistoScanning (™) . BJU Int 2012; 110: pp. 28-35.

  • 10. Braeckman J., Autier P., Soviany C., et. al.: The accuracy of transrectal ultrasonography supplemented with computer-aided ultrasonography for detecting small prostate cancers. BJU Int 2008; 102: pp. 1560-1565.

  • 11. Montironi R., van der Kwast T., Boccon-Gibod L., et. al.: Handling and pathology reporting of radical prostatectomy specimens. Eur Urol 2003; 44: pp. 626-636.

  • 12. Dickinson L., Ahmed H.U., Allen C., et. al.: Magnetic resonance imaging for the detection, localisation, and characterisation of prostate cancer: recommendations from a European consensus meeting. Eur Urol 2011; 59: pp. 477-494.

  • 13. Schatzkin A., Gail M.: The promise and peril of surrogate end points in cancer research. Nat Rev Cancer 2002; 2: pp. 19-27.

  • 14. McClure T.D., Margolis D.J., Reiter R.E., et. al.: Use of MR imaging to determine preservation of the neurovascular bundles at robotic-assisted laparoscopic prostatectomy. Radiology 2012; 262: pp. 874-883.

  • 15. Arumainayagam N., Ahmed H.U., Moore C.M., et. al.: Multiparametric MR imaging for detection of clinically significant prostate cancer: a validation cohort study with transperineal template prostate mapping as the reference standard. Radiology 2013; 268: pp. 761-769.

  • 16. Bubley G.J., Bloch B.N., Vazquez C., et. al.: Accuracy of endorectal magnetic resonance/transrectal ultrasound fusion for detection of prostate cancer during brachytherapy. Urology 2013; 81: pp. 1284-1289.

  • 17. Gangi A., Tsoumakidou G., Abdelli O., et. al.: Percutaneous MR-guided cryoablation of prostate cancer: initial experience. Eur Radiol 2012; 22: pp. 1829-1835.

  • 18. Niaf E., Rouviere O., Mege-Lechevallier F., et. al.: Computer-aided diagnosis of prostate cancer in the peripheral zone using multiparametric MRI. Phys Med Biol 2012; 57: pp. 3833-3851.

  • 19. Braeckman J., Autier P., Garbar C., et. al.: Computer-aided ultrasonography (HistoScanning): a novel technology for locating and characterizing prostate cancer. BJU Int 2008; 101: pp. 293-298.

  • 20. Tamada T., Sone T., Jo Y., et. al.: Prostate cancer: relationships between postbiopsy hemorrhage and tumor detectability at MR diagnosis. Radiology 2008; 248: pp. 531-539.

  • 21. Bloch B.N., Furman-Haran E., Helbich T.H., et. al.: Prostate cancer: accurate determination of extracapsular extension with high-spatial-resolution dynamic contrast-enhanced and T2-weighted MR imaging–initial results. Radiology 2007; 245: pp. 176-185.

  • 22. Selnaes K.M., Heerschap A., Jensen L.R., et. al.: Peripheral zone prostate cancer localization by multiparametric magnetic resonance at 3 T: unbiased cancer identification by matching to histopathology. Invest Radiol 2012; 47: pp. 624-633.

  • 23. Wu L.M., Xu J.R., Ye Y.Q., et. al.: The clinical value of diffusion-weighted imaging in combination with T2-weighted imaging in diagnosing prostate carcinoma: a systematic review and meta-analysis. AJR Am J Roentgenol 2012; 199: pp. 103-110.

  • 24. Mowatt G., Scotland G., Boachie C., et. al.: The diagnostic accuracy and cost-effectiveness of magnetic resonance spectroscopy and enhanced magnetic resonance imaging techniques in aiding the localisation of prostate abnormalities for biopsy: a systematic review and economic evaluation. Health Technol Assess 2013; 17: pp. vii-xix. 1–281

  • 25. Egevad L., Algaba F., Berney D.M., et. al.: Handling and reporting of radical prostatectomy specimens in Europe: a web-based survey by the European Network of Uropathology (ENUP). Histopathology 2008; 53: pp. 333-339.

  • 26. Riches S.F., Payne G.S., Morgan V.A., et. al.: MRI in the detection of prostate cancer: combined apparent diffusion coefficient, metabolite ratio, and vascular parameters. AJR Am J Roentgenol 2009; 193: pp. 1583-1591.

  • 27. Mena E., Turkbey B., Mani H., et. al.: 11C-Acetate PET/CT in localized prostate cancer: a study with MRI and histopathologic correlation. J Nucl Med 2012; 53: pp. 538-545.

  • 28. Shah V., Turkbey B., Mani H., et. al.: Decision support system for localizing prostate cancer based on multiparametric magnetic resonance imaging. Med Phys 2012; 39: pp. 4093-4103.

  • 29. Drew B., Jones E.C., Reinsberg S., et. al.: Device for sectioning prostatectomy specimens to facilitate comparison between histology and in vivo MRI. JMRI 2010; 32: pp. 992-996.

  • 30. Jhavar S.G., Fisher C., Jackson A., et. al.: Processing of radical prostatectomy specimens for correlation of data from histopathological, molecular biological, and radiological studies: a new whole organ technique. J clin pathol 2005; 58: pp. 504-508.

  • 31. Shah V., Pohida T., Turkbey B., et. al.: A method for correlating in vivo prostate magnetic resonance imaging and histopathology using individualized magnetic resonance-based molds. Rev Sci Instrum 2009; 80: pp. 104301.

  • 32. Jager G.J., Ruijter E.T., van de Kaa C.A., et. al.: Local staging of prostate cancer with endorectal MR imaging: correlation with histopathology. AJR Am J Roentgenol 1996; 166: pp. 845-852.

  • 33. Lencioni R., Menchi I., Paolicchi A., et. al.: Prediction of pathological tumor volume in clinically localized prostate cancer: value of endorectal coil magnetic resonance imaging. Magma 1997; 5: pp. 117-121.

  • 34. Sommer F.G., Nghiem H.V., Herfkens R., et. al.: Determining the volume of prostatic carcinoma: value of MR imaging with an external-array coil. AJR Am J Roentgenol 1993; 161: pp. 81-86.

  • 35. Schiebler M.L., Tomaszewski J.E., Bezzi M., et. al.: Prostatic carcinoma and benign prostatic hyperplasia: correlation of high-resolution MR and histopathologic findings. Radiology 1989; 172: pp. 131-137.

  • 36. Kimm S.Y., Tarin T.V., Lee J.H., et. al.: Methods for registration of magnetic resonance images of ex vivo prostate specimens with histology. JMRI 2012; 36: pp. 206-212.

  • 37. Hughes C., Rouviere O., Mege-Lechevallier F., et. al.: Robust alignment of prostate histology slices with quantified accuracy. IEEE Trans Biomed Eng 2013; 60: pp. 281-291.

Academic Radiology, Volume 21, Issue 8
Journals General Radiology
This post is licensed under CC BY 4.0 by the author.
Recently Updated
Contents